WorldScape3D: Fix crash with empty data directory outside the editor

[KazW]

Fixes #1531.

Problem

When a scene has a WorldScape3D node with an empty data_directory, running the game with the editor binary crashes with signal 11. This includes pressing Play in the editor. WorldScape3DData::initialize() calls select_data_directory() in every TOOLS_ENABLED build, and that does EditorNode::get_singleton()->add_child(...). A game run with the editor binary has no EditorNode, so the singleton is null.

Fix

• The folder dialog now only opens when Engine::is_editor_hint() is true (IS_EDITOR). Editor behaviour doesn't change.
• Anywhere else (a game run with the editor binary, or an export template), it logs WorldScape3D data directory not set. Terrain will not be loaded and continues. A game run from the editor shows this in Debugger > Errors.

Export templates never had the crash, because select_data_directory() isn't compiled into them. Before this change they silently skipped loading. Debug templates now print the same error. Release builds don't print it, because the module's LOG macro is compiled out without DEBUG_ENABLED.

Testing

On Linux (CachyOS), with a dev_build=yes editor built from this branch on master (96e43d3). I used the MRP from the issue, with a WorldScape3D node and an empty data_directory.

Run	master	This PR
Run the game, --headless --quit-after 60	signal 11 (2 out of 2 runs)	exit 0, error printed (3 out of 3 runs)
Headless editor, -e --quit-after 120	exit 0	exit 0
Same scene with data_directory set to an empty folder	exit 0	exit 0

I also tested it manually in the editor. Pressing Play runs the game, and the error appears in the Debugger's Errors tab.

Unrelated to this fix: with no regions loaded, a small raised square renders near the world origin on the flat background. It also happens on master and 26.3-stable when data_directory points to an empty folder. I'll look into it separately.

AI disclosure

Per AI_POLICY.md: I used Claude Code to find the cause, write the fix, run the builds and tests, and draft this description. The change was also put through an AI review before submission. I tested the change manually before submitting.

🤖 Generated with Claude Code

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Repository: Redot-Engine/redot-engine/.coderabbit.yaml
• Review profile: CHILL
• Plan: Advanced
• Run ID: 2afbe4ae-48d9-42d3-8efc-a2b49cf4a2a2

📥 Commits

Reviewing files that changed from the base of the PR and between 132958a and e24f032.

📒 Files selected for processing: 1

• modules/worldscape_3d/worldscape_3d_data.cpp

Included review availability: This review used your included allowance. Your plan provides up to 8 included reviews per hour; 6 remain after this review.

Walkthrough

When the data directory is empty, WorldScape3DData::initialize opens the directory selector only in the editor. In other contexts, it logs that terrain will not load.

Changes

WorldScape3D directory handling

Layer / File(s)	Summary
Empty-directory initialization modules/worldscape_3d/worldscape_3d_data.cpp	When the data directory is empty, the selector opens only when IS_EDITOR is true. Otherwise, initialization logs that terrain will not load.

Priority: ➖ Normal

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix · Severity of issue fixed: Medium

Suggested reviewers: olddev78, arctis-fireblight

Merge Risk: ⚪ Minimal · up to e24f0

The empty-directory case reports the problem without attempting terrain loading. No actionable merge risk is established.

Pre-merge checks: 5

Check name	Status	Explanation
Linked Issues check		The PR satisfies the coding requirements in #1531. For an empty data_directory, WorldScape3DData::initialize opens the directory picker only in editor mode. Other contexts log the required error a…
Out of Scope Changes check		The reported change is limited to empty-directory handling in modules/worldscape_3d/worldscape_3d_data.cpp. The editor-mode guard and non-editor error path directly implement #1531. No unrelated pro…
Docstring Coverage		Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 1 files.
Description Check		Check skipped - CodeRabbit’s high-level summary is enabled.
Title check		The title clearly and concisely describes the main change: preventing a crash when the WorldScape3D data directory is empty outside the editor.

Full details: Linked Issues check

Explanation

The PR satisfies the coding requirements in #1531. For an empty data_directory, WorldScape3DData::initialize opens the directory picker only in editor mode. Other contexts log the required error and continue without loading terrain data. The existing load path remains for configured directories. The reported headless and Play tests cover the requested execution modes.

Full details: Out of Scope Changes check

Explanation

The reported change is limited to empty-directory handling in modules/worldscape_3d/worldscape_3d_data.cpp. The editor-mode guard and non-editor error path directly implement #1531. No unrelated production changes are identified.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[KazW]

@OldDev78 since this is WorldScape3D related, can you take a look, please?

[OldDev78]

Nice, small patch. 👍
This bug was introduced when trying to automate the data directory selection, good catch.
We just need to simplify the error message.

[OldDev78]

LGTM

[DaveTheEggman]

LGTM